See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/228359282

Lateritic Soils for Dam Foundations and Dam Cores-Two Case Studies and
Their Typical Properties

Article

CITATIONS READS

2 6,242

5 authors, including:

Ronald Haselsteiner Cristian Osan

Björnsen Consulting Engineers Lahmeyer Hydroprojekt
53 PUBLICATIONS   80 CITATIONS    4 PUBLICATIONS   64 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

A dam project that is located at the south side of Turkey. View project

THE EMBANKMENT DAMS OF A FLOOD RETENTION BASIN IN POLAND View project

All content following this page was uploaded by Ronald Haselsteiner on 04 November 2014.

The user has requested enhancement of the downloaded file.

 Lateritic Soils for Dam Foundations and Dam Cores -Two Case
Studies and Their Typical Properties
A. Ronald Haselsteiner1, B. Hans-Georg Schuetz1, C. Cristian Osan1 and D. Björn Somdalen1
1
Fichtner GmbH & Co. KG, Renewable Energies & Environment, Sarweystraße 3, 70191 Stuttgart
E-mail: ronald.haselsteiner@fichtner.de
Abstract
The present paper discusses both the principal properties of
lateritic clay soils as foundation and construction materials,
and the derived conclusions for the pore water pressure
development in a rockfill dam with clay core on a residual
soil foundation. Special attention is given to the
consolidation behavior of a typical rockfill dam with a
lateritic clay core founded on lateritic clay foundation. The
data basis for this paper was obtained both from literature
review, and from site data collected for hydropower projects
in Laos and Uganda.
The aim of this paper is to provide a rough overview of some
important engineering aspects of lateritic clays, and to
present a simple method for the estimation of pore water
pressure dissipation if lateritic clays are used as dam core
material and/or foundation.
Introduction
Lateritic soils, or more precisely Ferralsols (FR) are usually
the product of an in-situ (lateritic) weathering process of a
basement rock, under tropical climate conditions. The genesis
results in typical lateritic weathering profiles, often having a
considerable clay layer on top. Lateritic soils are located in
tropical regions, where they are used as construction material
for embankment dams, roads, etc. In some cases the clay can
remain in the foundation, whereas in other cases the dam
engineer will require its removal prior to constructing the
dam. For example, rockfill dams will often require to be
founded on relatively strong rock. Usually, the lateritic clay
is a favorable core material for embankment dams. If a
concrete dam type, e.g. an RCC dam, is chosen, the clayey
overburden layers will usually have to be removed in order to
provide suitable foundation conditions. The excavation
should then, ideally reach unweathered rock formations.
In this paper, the typical geotechnical and engineering soil
properties of lateritic soils are described, based on literature
data and confirmed by soil investigation results from two
case studies. One of the case studies is for a hydropower
project in Uganda, where embankment dams are partly
placed on residual lateritic soils, and where the lateritic soil
in the form of clay, is also used for the core material. The
other case study is a project in Laos, where a soil
investigation program was carried out as part of a feasibility
study for a hydropower project. In the latter case the
abutments were covered by lateritic clays of considerable
thickness.
The particular in-situ conditions complicate the task of
deriving comprehensive data from laboratory testing. Both
relevant site experience and an extensive (critical) analysis of
test results have to be combined, in order to find applicable
soil parameters.
According to [20], in the case of lateritic soil several aspects
need to be questioned and discussed regarding the probability
of existing sink holes, the fabric and texture of the in-situ
weathered lateritic soil, the suitability as clay core fill
material and the in-situ permeabilities.
The Occurrence and Genesis of Lateritic Soils
Definition and Classification
Typical for the genesis of lateritic soils are the hot and wet
climatic conditions of the tropical and subtropical regions.
Lateritic soils are products of a three-stage in-situ weathering
and decomposition process [1] [24]. Their color is generally
reddish to yellowish, dependant on the water regime during
genesis, and on the mineralogical composition of the parent
rocks [2]. Their name is derived from “Ferrum” and
“Alumen” that are Latin for iron and aluminum. Usually the
in-situ weathered lateritic soil does not show any distinctive
stratification [3].
Another definition is given in [20]. “Lateritic soil” is the
traditional expression for soils defined and named as
Ferralsols [2], Ferrallitic soils [3] or Ferralites [5] according
to different soil classification systems and approaches [10].
Referring to the color, “Red Earth” or “Tropical Red Clays”
have also been used for the superficial soil that results from
the weathering of typical basement rocks, e.g. Granites or
Ultrabasites. Regionally, also the terms “Latosol”,
“Latossolo”, “Kaolisols” are still in use [2].
A uniform soil or rock nomenclature/classification system for
tropical residual soils does not exist. Consequently, in
literature, and particularly in engineering practice, different
classifications and nomenclature will be found. Even a mix-
up with similar soil types, e. g. Acrisols or Andosols, cannot
be excluded. As stated in [6] “It is unfortunate that the terms
‘lateritic clays’ and even ‘laterite’ are still used by some
engineers to describe any reddish tropical soil”. This supports
the thesis that not all “laterite” soils in literature were, in fact,
what they were assumed to be. The transition to other soil
types is gradual. Fortunately for engineering purposes it does
not matter whether the classification is correct, but that the
geological and engineering properties as predicted or derived
from testing are reliable. Ferralsols have several subclasses,
e. g. plinthic, red or yellow ferrallitic Latosols [5]. Some
sources propose “combined charts of Casagrande's plasticity
and Skempton's activity chart for identification of fine-
grained tropical soils” [18].
In the present paper the term “Lateritic Soil” is consistently
used, as it is commonly known and does not contradict any of
the various existing classification approaches. International
engineering practice treats lateritic soils under the topic
“Residual Soils” [3]. More precisely this paper addresses
particularly red clays, although lateritic tropical clays may
also exhibit a yellowish color which have similar properties
to the addressed red clays [11]. Andosols and Vertisols are
also residual soils which may have similar properties to red
clays and are discussed in [3]. The transition from a Ferralsol
(red clay) to an Andosol is smooth regarding their
engineering properties, their composition and their
occurrence. According to [6] “Red Tropical Clays” may be
Ferciallitic Andosols, Ferruginous soils, Ferrisols and
Ferralitic Soils classified corresponding to the soil and rock
classification system of the Geological Society Working
Party from the year 1990.
Ref [3] proposes a classification of residual soils based on a
consideration of their mineralogical influence. The problems
of using classical, existing classification systems are complex
and may be summarized as follows [3]:
1. Laboratory tests may be misleading due to the
particular importance of the in-situ structure of
lateritic soils.
2. Existing classification systems were not established
for tropical in-situ formed residual soils but mainly
for transported and remolded soils.
3. Existing correlations and empirical relationships
describing engineering parameters may not be valid
for residual soils.
In order to gain representative soil and rock samples the
method of sampling, the treatment of the samples and, of
course, the number and the origin of samples play an
important role in providing a reliable data basis for the
engineering properties [3] [6]. With increasing depth the
samples are less representative due to larger heterogeneity as
result of the in-situ weathering process [3].
Occurrence
In [2], [3] and [6] maps are given which show the worldwide
distribution of different soil types, particularly residual soils
including Ferralsols. Due to the climatic conditions which are
necessary for the typical genesis of recent lateritic soils, the
occurrence is concentrated close to the equator, ranging from
the southern part of South America to the mid of the United
States. In Africa the lateritic soils are mainly located from the
equator southwards. Asia, India and major parts of South-
East Asia, including Vietnam, Laos and Malaysia, are places
where residual and especially recent lateritic soils are
dominant [3]. Generally 20% of the recent tropical soils are
Ferralsols [4]. On old terrestric land masses, relic soils are
most likely to occur as Ferralsols which are considered to be
typical for perpetual wet climates. In Europe lateritic soils are
originated from the palaeozoic era and occur e.g. at the
Vogelsberg in Germany.
Genesis
In contrast to transported soils, the genesis of in-situ soils is
crucially affected by the climatic site conditions and the
topography, and by the nature of the parent rock itself. The
chemical (decomposition) and physical (disintegration)
weathering process is mainly summarized by the term
“lateritization”. In addition, biological weathering processes
can also have an effect on the soil genesis [3].
Particularly, the chemical reactions that are supported by all
types of weathering processes are responsible for the removal
of silica and bases, and for the concentration of iron and
aluminum oxides. These elements are responsible for the soil
color. The lateritic clays are most likely consisting of
Kaolonite or Halloysites, Smectites and Allophane at an
earlier weathering stage [5] [6]. Under humid climate
conditions a Kaolonite may undergo alteration to a Gibbsite.
The presence of cracks, fissures and joints in the parent rock
results in infiltration of water and transportation of dissolved
parts caused by chemical weathering. Local inhomogeneity
will also result in a corresponding weathering pattern.
Lateritic soils result from all types of rocks, either igneous,
metamorphic, or sedimentary rocks [24]. “Sandstones and
granites are the most common rocks subject to lateritization”
[3]. In literature this process is also called “ferralization”.
Typical Weathering Profile
According to international practice in rock engineering a
classification by six grades is usually applied for soil-rock-
profiles subject to lateritization (see Figure 1).
This paper focuses on the residual soils, as a result of the
lateritization process, which comprise zones IA and IB.
Below the residual soil, horizons IC, IIA and IIB are
composed of Saprolites and weathered rock formations. The
transition takes place gradually, which leads to an
inhomogeneous and indistinct stratification. Unweathered
rock forms horizon III and is simultaneously the parent rock
material for the genesis of the lateritic clay [3]. The parent
rock type is also important for the later engineering IC” or “V/VI” (Figure 1, [12], [13]) may reach a few meters,
properties of the soils. whereas the whole weathering profile frequently results in a
Corresponding to the lateritization process, the fabric and thickness exceeding 20m. Below the lateritic clay horizon
structure is usually dominated by the weathering process and further layers of moderately weathered and decomposed rock
genesis. For example, the laterite weathering profile in layers (Saprolites) are of varying thickness. The thickness
Uganda exhibited steeply inclined, almost vertical joints and can be similar to that of the soil horizon [4] or can exceed
fissures. even depths of 100m as reported for Saprolite thicknesses in
The soil and rock classification tends to be difficult to apply [5].
in practice, due to the gradual transition. Zones (I) and (VI)
are relatively easy to identify, but classification of the Properties and Engineering Parameters
intermediate zones is more problematic. The correct
classification of upper zones III to V is often important for General
the determination of an appropriate foundation depth. For Due to the variability of lateritic soils, laboratory and field
these strata, the assumed engineering design parameters will tests can only provide a general “average” of the real in-situ
have to be guaranteed which can only be assured by a properties. The determined values are generally more
competent and experienced site supervision. representative for the remolded and compacted soils used as
Classes W eathering DescriptionA) Fichtner (Uganda) – Classes
A)
ProfileA) ClassificationC) B) construction material, than for the specific local situation of
Description Project the residual soil. The description and definition of the
(Rock: Amphibolite) Classes
Top soil Influenced also by bioturbation and climate (changes). engineering parameters and properties, are as difficult as the
VI discoloured All rock material
Reddish Brown Clay IA definition, nomination and classification of lateritic soils [6].
variable lateritization
Residual rock converted to soilstructure
stage,
WVI “The main difficulty in dealing with these soils for
Soil and fabric destroyed
typical depth 0-4m engineering purpose is that their characteristics are very
Pale pink clay/silt, IB
ferruginous, relic
different from those of transported soils” [3]. This is also
All rock material
V decomposed and/or minerals and structures, based on the particular structure and fabric that differs from
disintegrated to soil. typical depth 4-20m
Compl. Original mass structure Pale greenish yellow
WV
IC
those of transported and/or remolded, compacted soils.
Weathered still largely intact silt/clay with fine sand, Due to the weak texture lateritic soils are subject to
decomposed rock relic miner. and struct.,
typical depth 20-30m
superficial erosion after topsoil stripping like other
>50% rock material II A unprotected soils. In this context, it is noted that Plinthosols,
IV decomposed and/or a subclass of Ferralsols, have hydromorphic properties and
disintegrated. Brownish grey, coarse
Highly
weathered
Fresh/discolored rock grained, weak strength therefore have to be treated with appropriate care if used for
present as discontinuous
framework or corestones foundation or for fill material.
<50% rock material
WIV/III In [5] the occurrence of “slickensides” in tropical residual
decomposed and/or soils is also mentioned. “Slickenside” surfaces are sliding
III disintegrated to soil. Grey, coarse grained,
Moderat. Fresh/discolored rock moderately strong/strong surfaces that generate from processes where high stresses
present, discontinuous
Weathered
framework or corestones
caused a rectification of the clay minerals. This results in
predefined slip surfaces within the soil, having effective
Discoloration due to Dark grey, coarse IIB
II
weathering of rock and grained, very strong, shear strength values significantly lower than the
Slightly discontinuity may weaker
weathered than fresh rock.
joint surfaces iron
stained
corresponding average shear strength of the soil complex.
IB Faintly Discoloration on major
WI/II
III In [6] it is explicitly stated that the color does not need to be
weath. discontinuity surfaces Dark grey, coarse
Fresh rock grained, very strong, a proper criterion for engineering classification since soils
IA No visible spot of rock
Fresh material weathering
joint surfaces fresh. with different colors may have the same engineering
Notes: properties. Although, the different soils may be correctly
A) Classification proposed by the Geological Society of London (Fookes, 1997)
B) Typical Weathering grades for Igneous and Metamorphic Rocks
classified according to soil and rock classification systems.
(Deere & Patton, 1971) Examples are Ferralsols (red, yellow) and Andosols (brown,
C) Highly weathered and moderately weathered are treated equally in terms of yellowish brown). The latter can also exhibit a reddish color.
engineering properties; this was also done for WI/II. Due to difficulties in engineering classification and a wide
variety of engineering parameters and properties,
Figure 1: Typical Lateritic Weathering Profile including Soil geotechnical and civil engineers frequently still consider the
and Rock Classification corresponding to [12] [13] [20] and “red lateritic clays” as “problem soils” and “troublesome”
the applied Uganda Project Classification when encountered on site [6]. However, an early awareness
may help to avoid mistakes and misinterpretation and may
lead to a correct and cautious handling of this “Red Earth”.
Lateritic soil profiles with a considerable depth (more than
20m) result from terrestrial weathering over long periods,
Case Studies: Laos and Uganda
even millions of years. The classified horizons for “IA, IB,
In Laos, Fichtner was contracted to evaluate the technical and
economical feasibility of a 120MW hydropower plant located
at a tributary to the Mekong River, close to the borders with
Myanmar and China, in the Northern part of the country. For
this study, a geotechnical field investigation campaign was
performed by a local contractor. Different dam types, e.g.
140m high rockfill and RCC dams, were considered at
different locations. All along the corresponding section of the
river, the potential abutments were covered by residual soils
with 40m to 60m thickness in the upper slope regions. The
soil could generally be identified as tropical lateritic clay.
The color typically changed from reddish to yellowish. At the
riverbed itself fresh rock was encountered. The parent rock is
an Andesite.
In Uganda, a 250MW hydropower project is under
construction on the Nile River. Fichtner is the Design
Engineer of the EPC contractor. The HEPP is a combined
structure consisting of three embankment dams, a gravity
dam, two different concrete spillway structures and a
powerhouse. All three embankment dams are zoned rockfill
dams with a total length of more than 500m and a maximum
height of approximately 35m. The rockfill dams are all
designed with a clay core. Both abutments and parts of the
foundation are covered by lateritic soils (clay) under which
Saprolite of more than 10m thickness is found. From the
project experience it can be stated that the problems related
to classification of the soils and their engineering properties
can be considered to be typical for lateritic soils. Although a
comprehensive field and laboratory investigation program
was carried out starting in the mid-1980s, the engineering
properties and design parameters of the rock foundation, of
the residual soil, and of the core fill material (residual clay
soil) left considerable room for interpretation. However, the
involved engineering parties representing the Owner, the
Design Engineer, the Contractor, the Check Consultants and
the Dam Safety Panel reached an agreement on the
engineering parameters after an extensive evaluation and
specific additional laboratory tests carried out in Europe. The
design parameters both allow a technically safe and
economical design. The parent rocks of the lateritic soil were
dominantly Amphibolites but also some schist (see also
Figure 1).
Index Properties and Consistency
Corresponding to statements in literature, e.g. [6] [24] [25],
lateritic or residual soils have a wide spread variability
regarding their engineering properties. This is also
documented by the liquid limit (LL) and the plasticity index
(PI). In Figure 2, several residual soils, mainly lateritic clays,
are presented referring to literature, applying literature data,
project archive data and data of the two recent case studies in
Laos and Uganda. According to [7] clays consisting mainly
of Kaolinites are generally located below the A-line in Figure
2. Also [20] states that lateritic soils are clays, sandy clays
and gravelly sandy clays which usually plot below the A-line.
The same is valid for clays consisting of Halloysites.
However, the given Casagrande diagram also shows that a
considerable number of soils with LL < 60% are located
above the A-line. The Andosols may plot in this area, too..
It should be noted that the accuracy of laboratory tests
depends crucially on the pretreatment of the specimen, and
on the testing procedure [3] [6]. Soil index test results are
typical and stand for the general variability of engineering
properties of lateritic clay soils. The lateritic clay as found at
the Uganda site is rather stiff having a cu-value (undrained,
unconsolidated) of cu = 50kPa. The index properties show
that this lateritic clay could be classified as CL to OL
according to USCS.
Figure 2: Casagrande diagram for various lateritic soils
(according to [3], [18] and [22])
In Laos, the investigated soil samples were limited to 24 tests
of remolded and undisturbed samples which indicate a
possible classification as CL according to USCS.
A detailed discussion of the soil indices of residual soils is
given in [3]. A-line plots for red clays are also given in [6],
differentiated by their parent rock types. Younger quaternary
parent rock types seem to generate rather plastic silts.
However, the parent rock type is only one aspect that
influences the genesis of lateritic soils, particularly of its clay
minerals.
Composition of Lateritic Soils
This paper contributes to a discussion of lateritic clay/silt
soils as one type of residual soil derived from different parent
rocks. Only the highly clayey strata of the lateritic profile are
included. Therefore, the soils for which the grading was
studied will always have a percentage of fines above 40% as
shown in Figure 3. In literature the range of investigated
lateritic soils varies from lateritic coarse gravel size, to high
plasticity clay as indicated by the data in Figure 2.
The soils encountered in Uganda and Laos had different
gradings as shown in Figure 3. In Laos a clayey-silty soil was
found, whereas in Uganda the lateritic soil has a relatively
wider grading. The grading diagram clearly shows two
ranges. However, a definite separation of the ranges was not
possible regarding both strata depth and sample site.
The occurrence of hard residual boulders [24] embedded in a For the design of embankment dams, the stress-deformation
relatively weak soil and rock matrix is also encountered in behavior, seepage conditions, and slope stability are
lateritic profiles. Evidently a significant number of boulders particularly important considerations. The shear strength
may provoke difficulties with regard to the subsoil treatment. parameters will be discussed in the next paragraph.
Attention has to be paid to this aspect since special As expected, the actual site data from the two case studies,
construction methods and/or the combination of different fall within limits typically found for lateritic soils.
techniques, e.g. jet grouting, cutoff walls and conventional The permeability of lateritic soils is dependant on several
grouting, may be necessary to reach a successful result. factors. However, it is clearly related to the weathering
process and its products. Under specific conditions pseudo-
sands may occur. These will naturally result in high
permeabilities, even if a corresponding clay fraction is
present. The lateritic soil genesis tends to form layers of high
permeability below the surface clay layer. This was observed
on-site in Uganda, where high water losses were seen both
during drilling and during in-situ water pressure tests. High
permeability layers, mostly located within the Saprolitic
subsoil area, have also been documented in [3] and [20]. In
general, the permeability of lateritic soils lies within a range
of k = 10-4m/s to 10-11m/s whereas [20] also mentions
maximum permeabilities of k = 10-2m/s in the Saprolite zones
where a vertical joint system usually prevails. Laboratory
tests of clays may also lead to lower values than encountered
Figure 3: Sieve curves for lateritic soils in Uganda, also referring to [16] for laboratory tests for a
clay core in Indonesia. In Laos the range of values was found
Other Geotechnical Parameters and Behaviour to vary from k = 10-4 to 10-7m/s. If lateritic clay is used as
In Table 1 selected engineering parameters for lateritic soils core material, the final in-situ permeabilities are one to three
are listed, together with a general range of values typically orders of magnitude higher than indicated by laboratory
found. The range of values found in the two case studies are testing. However, “typical values of permeability can be
also presented. problematic and misleading” [3] since local fabric may be
completely different from laboratory conditions. In [3] it is
TABLE 1: SELECTED SOIL AND ENGINEERING PARAMETERS OF also warned to give attention to the IC and IIA horizons in
LATERITIC SOILS respect to eventual piping development due to high
permeability and low stability in these strata.
General A few results of Oedemeter tests on the Uganda soils
Parameter Unit Laos Uganda
Range indicated a possibility of collapsibility when saturated.
Permeability Collapsible soils are usually found among deposit soils or in
m/s 10-2-10-11 10-4-10-7 10-4-10-9
kSat residual soils resulting from specific dry weathering
Permeability processes. Generally, for in-situ weathered soils a “lack of
Lu >2 4-30 2-30 information on the collapse mechanism” has to be stated [8].
Lugeon
Unit Weight In [8], a Brazilian Oxisol was investigated considering its
kN/m³ 14-22 15-18 16-22 collapsibility. The collapse potential of tropical residual soils
γ
Porosity is also discussed in [3]. In general, a low bulk density and a
% 25-60 45-54 26-46 low percentage of fines (percentage of particles retained by
n
No. 200 sieve) are crucial parameters for the collapsibility of
Liquid Limit
% >25 38-50 32-69 soils [9]. An interpretation of the findings in [17] leads to the
LL
conclusion that soils with a bulk density of more than
Plasticity Index
% 5-65 20-26 6-40 14kN/m³ and a percentage of at least 70% fines can be
PI
considered as not collapsible. The collapsibility can be
Fines Content % 10-100 50-90 13-98 classified by its collapse potential according to [9]. The
impact of collapse processes should be assessed in terms of
OWC % 10-35 20-32 14-32 their impact on the stability and durability of the structures.
For embankment dam structures, collapsibility in the
Depth of Bed. foundation may cause sink holes and/or considerable
[m] ~6-50 ~50 ~30
Rock (WIII-IV) settlements with subsequent cracking. In addition, a
corresponding piping process may be initiated. matric suction which is usually neglected providing a
Experience has shown that the residual moisture content of conservative design approach.
lateritic clays can be relatively high and may reach the The choice and determination of the shear strength
optimum moisture content (OMC), as determined in the parameters of lateritic soils, especially of impervious soils,
Proctor test. Therefore, the soil can often be used as fill may be a scientific work since laboratory and in-situ testing
material with minimal requirement for moisture conditioning are susceptible to be misleading and a time wasting process
(drying or watering). In many cases, however, the tropical [3]. Therefore, simple and reliable methods should prevail
dry climate and high precipitation will result in variable in- hand in hand with an appropriate design in order to cover all
situ moisture contents, and continual adjustment of the work possible uncertainties. Thus, the given shear strength values
process will be necessary, in order to achieve the required of lateritic clay soils are only shown to provide a general
moisture content. This is clearly of importance during indication.
construction, as a smooth and efficient construction process Usually, the 2/3-rule or the 1/2–rule are applicable methods
is necessary in order to achieve the required fill quantities, for the determination of reliable shear strength parameters
the required rate of construction, and the required quality. [23] in terms of preparing a safe and economical design. That
Investigations presented in [24] confirm that compaction on means that at least 2/3 or 1/2 of the test result plots should be
the wet side of OMC will most likely result in the lowest plotted above a selected design curve or values.
permeabilities, whereas compaction dry of optimum can Compacted clay cores usually exhibit stronger shear behavior
result in permeabilities of one magnitude higher. than residual in-situ soils. Residual lateritic soils in which the
structure of fracture planes still remains have in several cases
Shear Strength Parameters shown an agglomeration of certain minerals, e.g. Hematite, at
The determination of reliable design shear strength the upper part of former vertical joints. For existing planes
parameters has to consider the calculation tools, programs very low values for the friction angles ϕ’ = 5-7° (with c’ =
and methods applied for the stability analysis as well as the 0kN/m2) has been encountered also due to predefined slip
characteristics and properties of construction materials and surfaces with Hematite agglomeration or from planes
foundation layers, whilst paying attention to the predicted reported as “slickensides”.
stress range, stress-deformation behavior and the foreseen
degree of protection of the dam structure. Some values for Consolidation of Lateritic Soils Used for a
shear strength parameters are given in Table 2.
Zoned Rockfill Dam and Its Foundation
TABLE 2: SHEAR STRENGTH PARAMETERS OF LATERITIC
General
SOILS
The stability of embankment dams and related structures that
Soil Cohesion Friction Angle are founded on residual soils is affected by the development
Characteristic [kN/m²] [°] of excess pore water pressures since lateritic clays may
General Range exhibit impermeable, highly compressible properties.
Although, the typical fabric of in-situ weathered soils support
Peak c’ 0-25 ϕ’ 5-37
a fast and smooth pore water pressure dissipation by their
UgandaC)
joint system, excess pore water pressure will be encountered
Peak c’ 15 ϕ’ 25 in both compacted clay cores or in clay layers. The excess
Undrained cU 50B) ϕU 0A) pore water pressure behavior is particularly important for the
Residual cR 0A) ϕR 23 End of Construction (EoC), eventually also for the First
Laos Impoundment (FI), if impoundment takes place more or less
immediately after EoC and the consolidation process is still
Peak c’ 24 ϕ’ 17
going on, as well as for Rapid Reservoir Drawdown.
Undrained cU 30B) ϕU (18)A)B) Liquefaction during seismic events due to high pore water
Notes: pressures usually plays a subordinate role since lateritic clays
A)
Set to “Zero” due to safety aspects. have a considerable percentage of fines which withstand also
B)
Applying full excess pore water pressures, the undrained cohesion
critical pore water pressures, e.g. caused by dynamic loads.
cu and the undrained friction angle ϕu = 0° is one conservative
A combined 2D FEM seepage and stress-deformation model
method to evaluate the stability during consolidation phase.
C)
Due to safety aspects the same shear strength parameters were can be applied for modeling. However, in many cases the
applied both for the clay core and the residual soil, although efforts for the preparation of a corresponding model will not
higher values would have been justified for the clay core. be justified since it does not generate better or more reliable
results compared to simple analytical approaches because of
The strength behavior of compacted lateritic soils for both the varying work conditions and the inhomogeneity of
unsaturated conditions is discussed in [25]. Usually, shear the material. Thus, both the additional soil testing and dam
strength parameters increase in unsaturated soils due to monitoring in order to determine the real embankment
behavior for the later verification of the numerical model, do
not seem appropriate, eventually wasting both planning
resources and time. For small rockfill dams the deformation
behavior usually plays a subordinate role. Thus, the
application of a simplified 1D pore water pressure model was
considered to be justified for the Uganda project. Here, the
calculated excess pore water pressure during EoC and during
FI were input data for the stability analysis considering
effective stresses. The predicted pore water pressures were
considered by B-bar-values in the stability analysis according
to [19]. In Equation (1) the B-bar is defined.
Δu ⎡ Δσ ⎛ Δσ 3 ⎞⎤
B= = B ⋅ ⎢ 3 + A ⋅ ⎜⎜1 − ⎟⎟⎥ (1)
Δσ1 ⎣ Δσ1 ⎝ Δσ1 ⎠⎦
The B-bar-value [-] is derived from the quotient of the pore
water pressure changes Δu [kPa] and the changes of the
major principal stress Δσ1 [kPa]. The B-value reflects the
pore water pressure change Δu with regard to the change of
the minor principal stress Δσ3 [kPa] during triaxial testing.
For common saturated soils the B-value is usually equal to
1.0, consequently the B-bar-value is a function of the major
and minor principal stresses and the parameter A [-] which
considers the (over)consolidation condition according to
[27]. Several authors investigated the pore water pressure
behavior of soils and, therefore, gave hints for values of the
parameters A and B [13] [19] [20] [27].
Simplified 1D Approach
The simplified consolidation/pore water dissipation 1D-
model was applied in order to estimate potential excess pore
water pressures in the lateritic core and the residual soil
foundation. The applied approach is based on Terzaghi’s
consolidation theory [14] and on simplifications made for
typical rockfill dam designs as described in [15]. For the
performed estimation of the pore water pressure dissipation
Curve No. 1 (see [15]) was considered to be representative
for a zoned rockfill dam with a central clay core (Figure 4).
Figure 4: Pore water pressure dissipation/consolidation
curves (according to [14] and [15])
For this approach, it is necessary to know the consolidation
coefficient cV, the rate of construction, and the thickness and
number of compacted layers. Firstly, the drainage path length
and, finally, the excess pore water pressures are determined.
For this purpose the modeled layer thickness, the number of
modeled layers, and the time steps must be defined. The
approach of [15] postulates that the consolidation process
reacts similarly to the pore water pressure dissipation
behavior (Figure 4).
The rates of construction for the Uganda rockfill dams were
assumed to be between 0.15m/d and 0.30m/d as given in the
construction schedule. The consolidation coefficient values
ranged from cv = 10-7 to 10-9m²/s according to laboratory
tests. The thickness of the residual soil at the considered
critical sections varied from 5m to 20m, the core height was
assumed to reach a maximum of 35m and the core width a
maximum of 25m at the foundation.
Results
The results are shown in Figure 5 where the pore water
pressure coefficients are plotted versus the considered
consolidation coefficients. The results were approximately
similar for the core and for the foundation layer,
consequently it is applicable to summarize the results in one
plot. Curves were prepared for three different rates of
construction. The curves reflect the average values for
varying dam/core heights and foundation thicknesses.
Figure 5: Calculation example for the pore water pressure
coefficient (B-bar) versus the consolidation coefficient for
different rates of construction
The excess pore water pressures were implemented in the
stability analysis by constant B-bar-values for each modeled
soil zone. Since this is also a rough approximation of the real
pore water pressure conditions and this approach tends to
overestimate the real conditions, curves were multiplied by
2/3 assuming a parabolic pore water pressure distribution
along the slip surface.
The curves shown in Figure 5 were analyzed for EoC. The
applied model also enabled an estimation for further
dissipation after EoC and for the continuing consolidation
process. In this case, the pore water pressures in the relatively
thin foundation layer are likely to dissipate faster than in the
core. For the load case FI the pore water pressures in the
foundation are predicted to be already reduced to 10% of
maximum at EoC, whereas in the core still 40-80% of the
maximum calculated excess pore water pressures are still
effective.
Figure 5 shows also that almost full pore water pressures are
reached for consolidation coefficients cV < 10-7m²/s.
Otherwise, for cV > 10-4m²/s no remarkable excess pore water
pressures have to be expected. Therefore, the consolidation
coefficient derived from soil tests, may be one applicable
indicator as to whether significant excess pore water pressure
may occur within a dam structure under consideration.
Generally, it has to be noted “that it is difficult to predict the
pore pressure response with any degree of accuracy” [21] and
the described method and results can only reflect a rough
estimation of the actual in-situ conditions.
To underline the high variation of the pore water pressures
occurring in the foundation, Figure 6 shows an evaluation of
the pore water pressure coefficient for different lateritic clay
foundation thicknesses. The figure also considers different
embankment construction periods TCon for the load case End
of Construction. As expected, pore water pressures exhibit
higher values for shorter construction periods. Also, thin
subsoil layers < 2m encounter only maximum 30% of the
overburden stress in terms of applied conditions, particularly
the selected consolidation coefficient cV = 2.6 10-6m²/s.
Although, Figure 6 reflects a rough estimation, the deduced
curves should be universally valid for described boundary
constraints of typical embankment dams founded on
corresponding foundation soils.
Figure 6: Calculated pore water pressure coefficient (B-bar-
value) for residual soil layers of varying thicknesses beneath
embankment dams at End of Construction
Conclusion
Lateritic clay soils do have a wide range of engineering
parameters and accordingly show also a wide range regarding
their geotechnical behavior due to their specific composition
and genesis. Although this paper provides a brief overview of
typical engineering parameters and properties of lateritic clay
soils, engineers have to refrain from a generalized application
and from transferring results and interpretations from one to
another project, also if only little differences are expected.
Due to the observed and documented variation of parameters
and properties, a comprehensive investigation campaign
consisting of both field and laboratory tests must be carried
out in order to determine reliable design soil properties and
parameters.
Regarding the specific boundary constraints of the Uganda
project, average pore water pressure coefficients (B-bar-
values) were deduced from a simple analytical 1D approach.
B-bar-values of 40% to 65% for both the core and the
foundation layer were applied in terms of load cases
regarding unconsolidated states, e.g. End of Construction.
Although, site specific data was considered the deduced
results, particularly the curves in Figure 5 and Figure 6,
would seem to be universally applicable for approximate
estimations of pore water pressure development for similar
projects. Comparison with documented case studies [21] [26]
confirm this statement since the final results would seem to
be sufficiently accurate, and not “excessively” conservative.
Nevertheless, a project specific evaluation of the pore water
pressure development should be performed when the detailed
design and construction phases start.
In such cases, simple estimation methods generate
satisfactory results since the precise prediction of excess pore
water pressures is difficult and may be misleading due to the
sensitivity of the influencing factors, soil parameters and
local particularities. However, corresponding control
measurements should be performed during construction.
Although more and more embankment dams are built with,
and on, lateritic clays and the experience and knowledge has
increased considerably over the last decades, the tasks,
engineering aspects and requirements should not be
underestimated. Uncertainties in predicting excess pore water
pressures and effective shear strength values can have a
serious impact on the progress of a project, even if major
failure processes can be predicted and avoided during the
design and construction phase. With regard to the rate of
construction, the removal of impermeable foundation layers
which may be causing concern, may be preferable to
applying a sophisticated analysis and design, in order to
avoid critical excess pore water pressures, even at the
expense of slightly increased fill quantities. Lateritic clays
can be ideal clay core materials, exhibiting satisfactory shear
strength and low permeability characteristics. However, as
with all earthfill placement, water content must be carefully
controlled in order to ensure acceptable quality, whilst
achieving the required rates of construction.
Acknowledgements
Fichtner would like to thank all its Employers for the
permission to use the applied project data.
 References [23] USSD (2007): Strength of Materials for Embankment Dams. White
Paper, Committee on Materials for Embankment Dams, United States
Society of Dams (USSD), Denver (USA)
[1] Ganssen, R. (1965): Grundsätze der Bodenbildung. [24] Nwaiwu, C. M. O.; Osinubi, K. J.; Afolayan, J. O. (2005): Statistical
Hochschultaschenbücher, Bibliographisches Institut, Mannheim Evaluation of the Hydraulic Conductivity of Compacted Lateritic Soil.
[2] Glaser, B. (2005): Böden und Landnutzung in den humiden Tropen – Geotechnical Testing Journal, Vol. 28, No. 6, pp. 586-595
Ferralsole (FR). Department of Soil Physics, University Bayreuth [25] Mahalinga, U.; Williams, D. J. (1995): Unsaturated strength behaviour
[3] ICOLD (2008): Tropical Residual Soils As Dam Foundation and Fill of compacted lateritic soils. Geotechnique, vol. 45, no. 2, pp. 317-320
Material. ICOLD Committee on Material for Fill Dams, International [26] Li, C. Y. (1967): Construction Pore Pressures in Three Earth Dams.
Commission On Large Dams (ICOLD), ICOLD-Bulletin (Draft) Journal of the Soil Mechanics and Foundations Division, Proceedings
[4] Scheffer, F.; Schachtschnabel, P. (2002): Lehrbuch der Bodenkunde. of the American Society of Civil Engineers (ASCE), March 1967
15. Auflage, Spektrum Akademischer Verlag, Heidelberg [27] Skempton, A. W.; Bjerrum, L. (1957): A contribution to the settlement
[5] Kuntze, H.; Roeschmann, G.; Schwerdtfeger, G. (1994): Bodenkunde. analysis of foundations on clay. Geotechnique, vol. 7, no. 4, pp. 168-
5., neu bearbeitete und erweiterte Auflage, Verlag Eugen Ulmer, 178
Stuttgart
[6] Northmore, K. J.; Culshaw, M. G.; Hobbs, P. R. N.; Hallam, J. R.;
Entwisle, D. C. (1992): Engineering Geology of Tropical Red Clay
Soils – Summary Findings and Their Application for Engineering
Purposes. Technical Report WN/93/15, Overseas Development
Administration (ODA) & British Geological Survey (BGS),
Nottingham
[7] Day, R. W. (2000): Geotechnical Engineer’s Portable Handbook.
McGraw-Hill, New York
[8] Galvao, T. C. de Brito; Schulze, D. G. (1996): Mineralogical
Properties of a Collapsible Lateritic Soil from Mias Gerais, Brazil.
Soil Sci. Am. J., 60, pp. 1969–1978
[9] ASTM 5333-03: Standard Test method for Measurement of Collapse
Potential of Soils. American Society for Testing and Materials, Annual
Book of ASTM Standards 2007, Volume 04.08, Baltimore, USA
[10] Wesley, L. D.; Irfan, T. Y. (1997): Classification of Residual Soils.
Mechanics of Residual Soils. Editor: G. E. Blight, A.A., pp. 17-30
[11] Dumbleton, M. J.; Newill, D. (1962): A study of properties of 19
tropical clay soils and the relation of these properties with the
mineralogical constitution of these soils. British Road Research
Laboratory, Note 44
[12] Fookes, P. G. (1997): Tropical Residual Soils. The Geological Society,
London
[13] Deere, U.; Patton, F. D. (1971): Slope Stability in Residual Soils.
Proceedings 4th Pan-Am Conference Soil Mechanics Found. Eng. (San
Juan, Puerto Rico), vol. 1, pp. 87-170
[14] Terzaghi, K.; Peck, R. B. (1961): Die Bodenmechanik in der
Baupraxis. Springer-Verlag, Berlin Göttingen Heidelberg
[15] Narita, K.; Okumura, T.; Murata, N.; Ohne, Y. (1983): A simplified
method of estimating construction pore pressures in earth dams. Soils
and Foundations, Japanese Society of Soil Mechanics and Foundation
Engineering, vol. 23, no. 4, pp. 43-55
[16] Djarwadi, D.; Suhendro, D.; Hardiyatmo, H. C. (2004): Engineering
properties of kaolinitic clay as core as core of Batutegi dam. New
Developments in Dam Engineering. Editors: Wieland, Ren & Tan,
Taylor & Francis Group, London
[17] Mock, R. G.; Pawlak, S. L. (1983): Alluvial Fan Hazards at Glenwood
Springs. Geological Environment and Soil Properties, ASCE
Conference Proceedings, Houston, Texas, October 17-21, 1983, pp.
221-233.
[18] Vargas, M. (1982): Characterisation, identification and classification
of tropical soils. Proceedings of the 2nd International Conference on
geomechanics in Tropical Soils, Singapore, A.A. Balkema, Rotterdam,
vol. 2, pp. 469-487
[19] Skempton, A. W. (1954): The pore pressure coefficients, A and B.
Geotechnique, Vol. 4, pp. 143-147.
[20] Fell, R.; MacGregor, P.; Stapledon, D.; Bell, G. (2005): Geotechnical
Engineering of Dams. A. A. Balkema Publishers, Leiden London New
York Philadelphia Singapore
[21] Hunter, G.; Fell, R.; Khalili, N. (2000): The deformation Behaviour of
Embankments on Soft Ground. UNICIV Report No. R-391, University
of New South Wales, Sydney 2052 Australia
[22] Terzaghi, K. (1958): Design and performance of the Sasumua Dam.
Proceedings of the Institution of Civil Engineers, Harvard soil
mechanics series, no. 55, pp. 369-394

View publication stats